QVD: Post-training Quantization for Video Diffusion Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914874357972992 |
|---|---|
| author | Tian, Shilong Chen, Hong Lv, Chengtao Liu, Yu Guo, Jinyang Liu, Xianglong Li, Shengxi Yang, Hao Xie, Tao |
| author_facet | Tian, Shilong Chen, Hong Lv, Chengtao Liu, Yu Guo, Jinyang Liu, Xianglong Li, Shengxi Yang, Hao Xie, Tao |
| contents | Recently, video diffusion models (VDMs) have garnered significant attention due to their notable advancements in generating coherent and realistic video content. However, processing multiple frame features concurrently, coupled with the considerable model size, results in high latency and extensive memory consumption, hindering their broader application. Post-training quantization (PTQ) is an effective technique to reduce memory footprint and improve computational efficiency. Unlike image diffusion, we observe that the temporal features, which are integrated into all frame features, exhibit pronounced skewness. Furthermore, we investigate significant inter-channel disparities and asymmetries in the activation of video diffusion models, resulting in low coverage of quantization levels by individual channels and increasing the challenge of quantization. To address these issues, we introduce the first PTQ strategy tailored for video diffusion models, dubbed QVD. Specifically, we propose the High Temporal Discriminability Quantization (HTDQ) method, designed for temporal features, which retains the high discriminability of quantized features, providing precise temporal guidance for all video frames. In addition, we present the Scattered Channel Range Integration (SCRI) method which aims to improve the coverage of quantization levels across individual channels. Experimental validations across various models, datasets, and bit-width settings demonstrate the effectiveness of our QVD in terms of diverse metrics. In particular, we achieve near-lossless performance degradation on W8A8, outperforming the current methods by 205.12 in FVD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_11585 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | QVD: Post-training Quantization for Video Diffusion Models Tian, Shilong Chen, Hong Lv, Chengtao Liu, Yu Guo, Jinyang Liu, Xianglong Li, Shengxi Yang, Hao Xie, Tao Computer Vision and Pattern Recognition Artificial Intelligence Recently, video diffusion models (VDMs) have garnered significant attention due to their notable advancements in generating coherent and realistic video content. However, processing multiple frame features concurrently, coupled with the considerable model size, results in high latency and extensive memory consumption, hindering their broader application. Post-training quantization (PTQ) is an effective technique to reduce memory footprint and improve computational efficiency. Unlike image diffusion, we observe that the temporal features, which are integrated into all frame features, exhibit pronounced skewness. Furthermore, we investigate significant inter-channel disparities and asymmetries in the activation of video diffusion models, resulting in low coverage of quantization levels by individual channels and increasing the challenge of quantization. To address these issues, we introduce the first PTQ strategy tailored for video diffusion models, dubbed QVD. Specifically, we propose the High Temporal Discriminability Quantization (HTDQ) method, designed for temporal features, which retains the high discriminability of quantized features, providing precise temporal guidance for all video frames. In addition, we present the Scattered Channel Range Integration (SCRI) method which aims to improve the coverage of quantization levels across individual channels. Experimental validations across various models, datasets, and bit-width settings demonstrate the effectiveness of our QVD in terms of diverse metrics. In particular, we achieve near-lossless performance degradation on W8A8, outperforming the current methods by 205.12 in FVD. |
| title | QVD: Post-training Quantization for Video Diffusion Models |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2407.11585 |